Analysis of the turbulence data collected In Monterey Bay, California using the CIRPAS/NPS Twin Otter aircraft during the Autonomous Ocean Sampling Network (AOSN-II) project in 2003 suggested that bulk surface flux parameterizations and the commonly used surface similarity functions were not valid in many cases. The effect of sea surface waves, stable atmospheric stratifications due to cold sea surface temperature, coastal topography forcing, and stratocumulus cloud are factors that can lead to this discrepancy. In this work we analyze specific cases from AOSNII dataset in different wind directions (northwesterly, southeasterly, offshore and weakly onshore). Surface turbulent fluxes (sensible, latent, and momentum fluxes) were obtained from straight flight legs with the eddy correlation method. In addition to turbulent fluxes, the divergence and curl of wind and wind stress were estimated after an interpolation of surface momentum fluxes on a regular grid with 5 km spacing. The flights were carried out from the morning to the early afternoon of each day. We also use in the analysis wave and sea surface current data from nearby buoys and HF radars and try to connect the observed discrepancies in the flux parameterization with sea state variability and cloud dynamics. Finally, comparison of observed surface flux fields with the Navy COAMPS mesoscale model will be presented to understand interaction of the various components affecting coastal surface fluxes.